Analysis of Lossless Compressors Applied to Integer and Floating-Point Astronomical Data

Ôscar Maireles-González*, Joan Bartrina-Rapesta, Miguel Hernández-Cabronero, Joan Serra-Sagristà

*Corresponding author for this work

Research output: Chapter in BookChapterResearchpeer-review

1 Citation (Scopus)

Abstract

In this work, lossless compression algorithms are evaluated on a variety of real, current as-tronomical images. The test dataset comprises raw (integer) and processed (floating-point) images of discrete and extensive astronomical objects, captured by spatial or terrestrial tele-scopes. Compression techniques herein analyzed are chosen to be representative of the most recent algorithms devised for astronomical data, as well as the most commonly employed compressors employed in real observatories. Experimental results suggest that coding techniques such as RICE and HCOMPRESS, typically employed in world-class observatories such as Roque de los Muchachos, do not produce the best possible lossless compression results. Instead, JPEG-LS, LZMA and NDZIP yield the best compression ratio results for 16-bit data (2.72), floating-point data (2.38) and radio data (1.81), respectively. Therefore, the efficiency with which data are stored and transmitted by these observatories could be significantly improved by selectively employing the aforementioned algorithms.

Original languageEnglish
Title of host publicationProceedings - DCC 2022
Subtitle of host publication2022 Data Compression Conference
EditorsAli Bilgin, Michael W. Marcellin, Joan Serra-Sagrista, James A. Storer
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages389-398
Number of pages10
ISBN (Electronic)9781665478939
DOIs
Publication statusPublished - 2022

Publication series

NameData Compression Conference Proceedings
Volume2022-March
ISSN (Print)1068-0314

Fingerprint

Dive into the research topics of 'Analysis of Lossless Compressors Applied to Integer and Floating-Point Astronomical Data'. Together they form a unique fingerprint.

Cite this